Intermediaries in the Medicare Market

Last registered on March 01, 2026
View Trial History

Pre-Trial

Trial Information

General Information

Title
Intermediaries in the Medicare Market
RCT ID
AEARCTR-0014329
Initial registration date
September 11, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
September 17, 2024, 11:51 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
March 01, 2026, 12:36 PM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Elaine Shen
Affiliation
UC Berkeley
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Margaret Kallus
PI Affiliation
UC Berkeley
Contact Investigator

Additional Trial Information

Status
In development
Start date
2024-10-16
End date
2026-10-15
Keywords
Behavior, Finance & Microfinance, Health, Labor
Additional Keywords
Medicare, Health Insurance, Brokers, Agents, Intermediaries
JEL code(s)
I13, I18, D14, D18, G41
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study seeks to understand the role of intermediaries in the Medicare market. Specifically, we investigate how quality of insurance advice is correlated with consumer and agent characteristics.
External Link(s)

Registration Citation

Citation
Shen, Elaine and Margaret Kallus. 2026. "Intermediaries in the Medicare Market." AEA RCT Registry. March 01. https://doi.org/10.1257/rct.14329-6.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2024-10-16
Intervention End Date
2026-10-15

Primary Outcomes

Primary Outcomes (end points)
The key outcome of interest is the quality of the Medicare plan recommendations from the insurance producer (or insurance broker or agent). We will examine two main outcomes: i) whether or not the agent recommends Medicare Advantage (MA), ii) the difference between the recommended Medicare Supplement (Medigap) plan premium and the lowest-cost Medigap plan premium available to the consumer. For simplicity, we also create a "bad recommendations" index combining these two outcomes into a single metric. We also measure the Medigap outcome two additional ways: iii) the Medigap Agent Markup (the difference between the recommended Medigap premium and the lowest Medigap premium offered by insurance companies the agent is currently appointed with) and iv) the Medigap percent markup (the Medigap Markup as a percent of the price of the lowest-cost identical plan. More specifics on how we measure the primary outcomes are included in our pre-analysis plan.
Primary Outcomes (explanation)
These are the primary outcomes because they are the most direct measures of recommendation quality for our scenario. Whether Medicare Advantage or Original Medicare is the correct recommendation depends on the specific scenario and senior characteristics.

Secondary Outcomes

Secondary Outcomes (end points)
We will look at whether agents use certain sales tactics and accurately explain important insurance concepts to seniors. More specifics on this are included in our pre-analysis plan.
Secondary Outcomes (explanation)
These outcomes are secondary because they seek to explain mechanisms or are less direct measures of quality.

Experimental Design

Experimental Design
We will conduct a correspondence study that randomly assigns seniors to call insurance agents for a plan recommendation. We will evaluate whether recommendation quality varies with caller and agent characteristics. Finally, we evaluate whether agent advice is elastic to a randomly assigned “competition” treatment and to different commission/incentive regimes.
Experimental Design Details
Not available
Randomization Method
Insurance agents are randomly assigned to treatment or control status by a computer. Consumers are randomly assigned by the secret shopper firm.
Randomization Unit
Consumers will be randomly assigned to insurance agents at the individual level. Variation in competition signaling and consumer characteristics such as gender and age will also be randomly assigned.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
With our current research funding, we expect to be able to collect between 650-950 observations. For additional detail, please see our pre-analysis plan.
Sample size: planned number of observations
With our current research funding, we expect to be able to collect between 650-950 observations. For additional detail, please see our pre-analysis plan.
Sample size (or number of clusters) by treatment arms
We will aim to have an equal number of observations in our treatment and control arms.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We are powered to detect differences of 10% in the Medicare Advantage recommendation rate and differences of $75 per year in annual Medigap premium recommendations.
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
UC Berkeley Committee for the Protection of Human Subjects
IRB Approval Date
2024-05-22
IRB Approval Number
2024-04-17373
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information